| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1999;68:1029-1033
© 1999 The Society of Thoracic Surgeons
a Division of Cardiothoracic Surgery, Southern Illinois University School of Medicine, Springfield, Illinois, USA
Address reprint requests to Ms Boley, Southern Illinois University School of Medicine, PO Box 19638, Springfield, IL 62794-9638
e-mail: tboley{at}siumed.edu
Abstract
Background. Lung volume reduction operations have been shown to improve the quality of life and functional status of some patients with end-stage emphysema.
Methods. Because of a perceived increase in the occurrence of postoperative gastrointestinal (GI) complications, we reviewed our experience in 287 patients who had lung volume reduction operations to determine the frequency of GI complications and to identify risk factors.
Results. Using a broad definition of postoperative GI complications (nausea, vomiting, abdominal distension, gastroesophageal reflux, diarrhea, constipation) there were 137 complications in 67 patients (23%). More severe GI complications (bowel ischemia, GI bleeding, perforation, ulceration, ileus, colitis, cholecystitis, pancreatitis) occurred 49 times in 27 patients (9.4%). Seven of the 27 patients required abdominal operations. Risk factors identified as predictive of severe complications include diabetes (p = 0.0003), lower preoperative hematocrit (p = 0.01), steroid use (p = 0.02), and use of parenteral meperidine analgesic (p = 0.002). Stepwise logistic regression demonstrated that diabetes was 7.02 times more likely to produce severe complications. Other risk factors included steroids (2.81), number of different pain medications (2.59), hematocrit decrease of 5% (1.96), and hematocrit decrease of 1% (1.14). In the patients with severe GI complications there were six of 27 (22%) hospital deaths compared with five of 260 (2%) in those without GI complications (p = 0.0001).
Conclusions. Severe GI complications in patients with emphysema who had lung volume reduction operations are not uncommon (9.4%) and influence the perioperative mortality rate. Heightened awareness to identified risk factors will allow earlier recognition, prevention, and perhaps decrease morbidity and mortality rates in these high-risk patients.
Emphysema affects 13.5 million people and is the fastest growing cause of mortality and morbidity in the United States [1]. In the past 5 years there has been an increased enthusiasm for surgical treatment of a subset of patients with emphysema. The efficacy of reduction pneumoplasty has been demonstrated by improvements in functional status, exercise tolerance, pulmonary function, and alveolar gas exchange [2–4]. The operation itself carries significant morbidity, mortality, and cost; therefore, surgeons have tried to optimize patient selection and the results of this operation [2, 5].
The typical emphysematous patient has chronic hypoxia as well as other associated medical problems. Long-term cigarette smoking leads to problems such as ischemic heart disease, diabetes, peripheral vascular disease, and gastrointestinal (GI) problems. Advanced age, chronic hypoxia, and steroid use might also be critical factors that contribute to the multiple problems in the emphysematous population. Serious GI complications have occurred in patients after lung volume reduction surgery (LVRS) and might be caused by the deleterious effects of chronic hypoxia as well as the use of corticosteroids [6–8]. These GI problems might increase mortality and morbidity. By identifying predictive factors for GI complications during the perioperative period, we might improve the results of this operation. This retrospective study was designed to determine the incidence of GI complications in patients who had LVRS and then to identify the important preoperative and perioperative risk factors.
Patients and methods
A retrospective chart review was conducted on 287 patients who had LVRS to identify the frequency of GI complications concomitant with hospitalization for LVRS as well as potential risk factors.
The following variables were identified as possible preoperative risk factors: age, sex, hypertension, atrial fibrillation, coronary artery disease, peripheral vascular disease, diabetes mellitus, obesity, GI history, exercise capacity, pulmonary function test results, hematocrit or hemoglobin level, blood gases, and preoperative medications. Four postoperative risk factors were examined, including type of operation, postoperative medications (corticosteroids, pain reducing medications, bronchodilators, cardiovascular drugs, and psychotropic or anxiolytic drugs), duration of air leak, and time to resumption of oral feeding. Hypertension was defined as taking antihypertensive medication. Coronary artery disease was defined as previous coronary artery bypass grafting, percutaneous transluminal coronary angioplasty, myocardial infarction, or positive angiography. Peripheral vascular disease was defined as previous revascularization or positive results of Doppler studies. Diabetes mellitus was defined as taking oral hypoglycemic medication or insulin dependence. Obesity was defined as 1
times ideal body weight. Gastrointestinal history included peptic ulcer disease, pancreatitis, GI cancer, diverticulitis, GI bleeding, inflammatory bowel disease, and cholecystitis. Preoperative medication included antiarrhythmics, antihypertensives, diuretics, oral hypoglycemics, nonsteroidal antiinflammatory drugs, corticosteroids, H2 blockers, nitrates, laxatives, and aspirin. The components of blood gases included pH, partial pressures of oxygen and carbon dioxide, and percent saturation.
Postoperative risk factors were medications or events before the development of GI complications. Minor complications were determined by patient complaint and confirmed by health care provider’s assessment. Major complications were confirmed by laboratory analysis, endoscopy, radiography, or intraoperative findings.
Minor GI complications included distension, constipation, nausea and vomiting, reflux, and diarrhea. Minor complications were able to be treated conservatively with observation or pharmacology. Major GI complications included bowel ischemia, GI perforation, pancreatitis, colitis, GI bleeding, ulceration, ileus, and cholecystitis. Major complications required pharmacology and some invasive procedures ranging from nasogastric tube insertion to abdominal operations.
Statistical analyses included 
2 test of independence and Fisher’s exact test to examine categoric variables as indicated by sample size. Continuous variables were analyzed using Student’s t test. Stepwise logistic regression was used to determine which variables provided the best predictors of complications. Variables that were significant by univariate analysis at p less than 0.05 were used in the stepwise logistic regression.
Results
During a 3-year period, 200 (77%) men and 87 (23%) women had LVRS for emphysema. The mean age of the patients was 66.7 years (range, 41 to 84 years). Patients had stapled reduction with a unilateral thoracoscopic (n = 257, 89.5%), bilateral thoracoscopic (n = 5, 1.7%), or median sternotomy (n = 25, 8.7%) approach. There were 137 minor and major GI complications in 67 patients (23.3%). The occurrence of GI complications did not appear to be related to the learning curve, as the 137 complications occurred sporadically during the 3 years. Forty patients (13.9%) had minor and 27 (9.4%) patients had major complications. Seven (26%) of the 27 patients with major complications required abdominal operations. Procedures included one abdominal exploration, five bowel resections, and four colostomies for necrotic bowel, one gastric ulcer repair, and one cholecystectomy.
Corticosteroids, diabetes, oral narcotic pain medications, atrial fibrillation, intravenous or intramuscular meperidine, and the duration of the chest tube were related to the development of minor GI complications. The use of prophylactic gastrointestinal protection (stool softeners, coating agents, or H2 blockers) was not related to a decrease in GI complications. Predictors of major complications included diabetes, lower preoperative hematocrit, intravenous or intramuscular meperidine, use of corticosteroids, and the use of multiple pain medications (Table 1). Diabetes was monitored by glucose checks four times per day. Supplemental insulin was administered as necessary to control glucose levels until patients could return to their routine preoperative insulin or oral hypoglycemic agents. Patients without complications had a mean preoperative hematocrit of 43.2 versus 40.9 in those with complications. Eleven of 27 (41%) patients who received meperidine had serious GI complications, whereas 16 of 226 (7.0%) patients who did not receive meperidine had complications. Corticosteroids were used in 18 of 26 (69.2%) patients with complications compared with 13 of 259 (43.6%) without complications. Although individual pain medications other than meperidine did not increase GI risks, the use of multiple types of analgesic drugs was related to the development of serious complications (2.9 different analgesic medications in patients without complications versus 3.6 in patients with complications).
|
|
|
Comment
Postoperative gastrointestinal complications have been noted after other thoracic operations, and etiologic factors have been defined [9–14]. The rate of GI complications after cardiopulmonary bypass is 0.3% to 2%, and the mortality rate after major GI complications in open heart operations patients is 11% to 59% [12]. Smith and associates [13] found a rate of GI complications in lung transplant patients of 16%, accounting for 22% of the deaths in lung transplantation patients. They stressed the need for early recognition of GI problems to reduce the mortality rate.
In this study, we attempted to determine the frequency of GI complications and to identify associated risk factors in patients who have LVRS. Gastrointestinal complications were divided into two subgroups, minor and major. Minor complications did not increase the mortality rate; however, major complications occurred in 9.4% of the patients and the mortality rate was increased (22%). This mortality rate constitutes 54% of the 3.8% mortality rate in our entire LVRS patient population.
Three risk factors were determinants of both minor and major GI complications (diabetes, corticosteroid use, and a higher number of pain medications). These risk factors indicate that minor GI problems are predecessors for major GI complications. Although minor GI complications did not contribute to the mortality rate, they could serve as a hazard for the development of major complications. Therefore, simple GI disturbances after LVRS should not be neglected and should be treated promptly.
Diabetes mellitus was the most important risk factor for minor and major complications after LVRS. In comparison, diabetes has not been found to be a risk factor for GI problems in open heart operations. Chronic hypoxia has been shown to render the GI system more vulnerable by elevating the gastric pH as well as the interstitial pH of the entire GI tract [7, 15]. The interstitial pH of the GI system is the most sensitive aspect of adequate tissue oxygenation [15]. In the present study, the amount of hypoxia in arterial blood gases was not predictive of postoperative GI outcome. Chronic diabetes mellitus affects the microcirculation of all tissues as well as the GI system. This defective microcirculation related to diabetes mellitus might be the precipitating factor for GI events on an already vulnerable GI system affected by chronic hypoxia.
The association between long-term corticosteroid use and gastrointestinal perforation is well established. The reports of perforation of multiple sites of the GI tract in patients who received corticosteroids has been reported repeatedly [16–21]. ReMine and coworkers [6] demonstrated that early diagnosis and surgical intervention after colonic perforation resulting from corticosteroid use reduced the mortality rate. High doses of corticosteroids could mask the symptoms of peritonitis and GI perforation [6]. In our study, major complications were 2.81 times more likely to occur in patients with corticosteroid treatment after LVRS, but these events did not appear to be dose related. All patients taking corticosteroids should be monitored closely for GI symptoms after LVRS.
Another risk factor identified by multivariate analysis was number of pain medications. A variety of postoperative pain medications, including meperidine, morphine, propoxyphene and acetaminophen, oxycodone and acetaminophen, acetaminophen and codeine, ketorolac, and aspirin, were used. Meperidine was the only individual drug that was a risk factor for GI complications by univariate analysis. This result was the case although other narcotics, including morphine and propoxyphene and acetaminophen, were used more commonly. However, an increasing number of different types of pain medication were associated with minor and major GI complications by both univariate and multivariate analysis. We believe that among these pain medications, meperidine should be used cautiously and multiple medications for postoperative pain management should be avoided.
In our study, nonsteroidal antiinflammatory drugs were not individually predictive of GI complications. However, patients were given only ketorolac for the first 48 postoperative hours, and they were avoided in 14 patients who had a history of GI or renal disease. Conventional nonsteroidal antiinflammatory drugs inhibit the synthesis of cytoprotective prostaglandins by inhibiting the cyclooxygenase 1 in the GI tract. They also cause arachidonic acid accumulation (precursor of prostaglandins and leukotrienes) and enhance the generation of leukotrienes, which causes neutrophilic adhesion to the endothelium resulting in vasoconstriction and tissue hypoxia. Another adverse effect of nonsteroidal antiinflammatory drugs is the effect on carboxyl groups, which causes inhibition of oxidative phosphorylation and lowers adenosine triphosphate generation, resulting in a loss of mucosal integrity and increased mucosal permeability [22].
Narcotic analgesics cause respiratory and GI depression. Because of these adverse reactions pain medications should be selected carefully, particularly in this subset of patients who are predisposed to respiratory depression and GI complications. One way to reduce pain medications might be to examine the surgical approach. Some surgeons have routinely used a thoracoscopic approach to reduce postoperative pain. Pain-related morbidity among thoracoscopy, muscle-sparing thoracotomy, and standard posterolateral thoracotomy has been well documented [23–25]. Although both thoracoscopy and median sternotomy were well tolerated in our patient population, no prospective studies are available presently to identify the least painful surgical approach.
The hematocrit value was also a risk factor for major GI complications. Unnecessary blood transfusions are not recommended to avoid GI complications, but GI complaints in patients with a lower hematocrit value should be investigated carefully. These patients could require blood transfusions to keep adequate tissue oxygenation because of their chronic obstructive pulmonary disease. We recommend consideration of maintaining a minimum hematocrit near 30% in LVRS patients.
Other factors that were not risk factors for GI complications after LVRS included GI history, age, sex, exercise capacity, systemic atherosclerosis, results of pulmonary function tests, and hypertension. The effects of mechanical ventilation or positive pressure ventilation could not be determined in this study because no patient required mechanical ventilation after completion of LVRS. Chest tube duration and atrial fibrillation were risk factors for minor complications by univariate analysis. By increasing the size of the patient population additional risk factors might be identified.
Reports of LVRS have focused on the relationship between pulmonary parameters and death [2, 3, 26]. The most common cause of death after LVRS is multiple system organ failure accompanying respiratory failure. The triggering event for multiple system organ failure could be unrelated to the respiratory system. After the initial event occurs, respiratory failure routinely follows because of the patient’s presenting diagnosis. The trigger site for multiple system organ failure was the GI system in 54% of the deaths in the present study.
Thoracic surgeons are well aware that LVRS improves the quality of life and functional status of carefully selected patients with end-stage emphysema recalcitrant to medical treatment [27, 28]. The current focus with LVRS is to continue to improve outcome. We identified several risk factors that predispose LVRS patients to GI complications. Diabetes is the greatest risk factor and is generally not modifiable in this population of patients. However, some risks can be reduced by minimizing corticosteroid and analgesic use and maintaining adequate hematocrit levels. Awareness should be heightened to the identified risk factors to allow earlier recognition, prevention, and perhaps to decrease morbidity and mortality rates in this high-risk subset of patients.
References
This article has been cited by other articles:
|
|
 |
|
  E. A. Hessel II Abdominal Organ Injury After Cardiac Surgery Seminars in Cardiothoracic and Vascular Anesthesia, September 1, 2004; 8(3): 243 - 263. [Abstract] [PDF]
|
|
|
|
 |
|
 I. Sanisoglu, M. Guden, Z. Bayramoglu, E. Sagbas, C. Dibekoglu, S. Y. Sanisoglu, and B. Akpinar Does off-pump CABG reduce gastrointestinal complications? Ann. Thorac. Surg., February 1, 2004; 77(2): 619 - 625. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. R. Stirling, W. J. Babidge, M. J. Peacock, J. A. Smith, K. S. Matar, G. I. Snell, D. J. Colville, and G. J. Maddern Lung volume reduction surgery in emphysema: a systematic review Ann. Thorac. Surg., August 1, 2001; 72(2): 641 - 648. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
